Carcinogenesis vol.13 no. 10 pp. 1763-1768, 1992
Analysis of residual amino acid—DNA crosslinks induced in intact cells by nickel and chromium compounds
Xinhua Lin, Zhbdong Zhnang and Max Costa1 New York University Medical Center, Nelson Institute of Environmental Medicine, New York, NY 10016, USA 'To whom all correspondence should be addressed
Introduction DNA-protein crosslinks represent an important lesion induced by chromium (Cr) and nickel (Ni) compounds in intact cells (1,2). These lesions are relatively persistent and not readily repaired (3,4). Previous studies have focused on examining the proteins complexed to DNA by Cr and Ni compounds in intact cells (2,5 - 7 ) . Cr has been shown to crosslink actin and other proteins found in the nuclear matrix while Ni compounds complexed both non-histone proteins and possibly histone HI (6,8). However, the complexing of proteins by Ni compounds in general has been found to be disrupted following treatment with SDS, while chromate-induced DNA—protein crosslinks were generally resistent to SDS treatment (5,9). Analysis of proteins complexed with DNA is plagued by problems of background inherent in DNA isolation procedures without proteases. Cr(in) was believed to participate in a substantial proportion of the Cr-induced DNA—protein crosslinks, whereas the binding of Ni 2+ to DNA and proteins was not as stable as the binding involving Cr(III) (6—9). Recent evidence has implicated both chromate and Ni compounds in the induction of oxidative DNA damage, and the possibility exists that the DNA-protein
Materials and methods Cell culture and amino acid labeling CHO cells were seeded at 4 x 10* cells in a 150 mm dish in 20 ml of complete medium containing 10% fetal bovine serum in an atmosphere of 5% CO}. After 4 h, radioactive amino acids (all were [3H] except cysonol meth [MS] or threonine [I4C]) were added at 10 jiCi/ml, and cells were incubated for another 24 h [cysteine 1105.7 (Ci/mmol), methionine 1162.9, tryptophane 17.9, glutarruc acid 25.0, aspartic acid 22.0, histidine 51.5, alanine 85.0, arginine 45.0, isoleucine 111.7, leucine60.0, rysine 85.7, phenylalanine 49.7, proline 35.8, valine 71.0, serine31.1, threonine 0.233, tyrosine 50.0]. Media deficient in each of the respective amino acids was utilized during labeling to enhance the sensitivity of detecting the amino acid-DNA complex. Cells were then washed twice with normal saline and 20 ml of complete a-MEM medium was added. Cells were exposed to various concentrations of NiC^ or K2CrO4 for 20 h as indicated in the figures and tables. Following this exposure, cells were scraped with a rubber policeman and collected by centrifugation at 1500 r.p.m. for 5 min. Cells were washed twice gently with PBS and resuspended in 10 ml of mis buffer. A sample of this suspension was withdrawn for determining the specific activity of the cellular proteins. The remaining cells were collected by centrifugation and placed in a digestion buffer containing 100 mM sodium chloride, 10 mM Tris, pH 8.0 and 0.5% SDS. This suspension was vortexed and allowed to stand to room temperature for 10 min. RNase (10 jig/ml) was added and the suspension was incubated at 37°C for 30 min. Proteinase K was subsequently added to a final concentration of 500 /jg/ml and the sample was digested at room temperature with gentle shaking overnight. The digested sample was thoroughly extracted twice with equal volume of phenol/chloroform/isoamyl alcohol (25:24:1) followed by another extraction with chloroform/isoamyl alcohol (24:1). The aqueous layer was transferred to new tubes and l/20th vol of saturated ammonium acetate and 2 vol of 100% ethanol were added and the mixture was allowed to stand at - 2 0 ° C for 1 h.
1763
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
Chinese hamster ovary cells were incubated with radioactive amino acids, the DNA was isolated by standard proteinase K/phenol/chJoroform extraction and residual amino acids complexed to the DNA were examined as an index of metal induced DNA-protein crosslinks. Using this method, both chromate and nickel caused residual histidine and cysteine to be complexed with the DNA isolated from metal-treated cells. In the case of chromate, a number of ammo acids were studied and Tyr, Thr and Cys were found to be complexed to DNA at a level (above the untreated control) that was statistically significant. Stability studies indicated that some of the chromate-induced DNA-protein complexes were mediated by direct participation of chromium (HI), whereas others that were resistent to dissociation by EDTA and mercaptoethanol did not seem to involve direct chromium(III) participation. A significant portion of the cysteine complexed to DNA by chromate was believed to involve glutathione since treatment of cells with cycloheximide did not decrease chromate-induced cysteine-DNA crosslinks. In the case of nickel, most of the stable DNA—protein crosslinks did not involve direct metal participation and were probably oxidatively mediated by Ni(II)/Ni(III) redox cycling. These findings present new methodology for analysis of DNA-protein crosslinks by examination of residual ammo acids associated with the DNA. This method should be highly sensitive and will yield important information about the mechanism of metal-induced DNA-protein crosslinks.
crosslinks induced by some of these compounds may be oxidatively mediated (10). Thus, while Cr(IH) may participate in some DNA-protein crosslinks, additional crosslinks induced by chromate in intact cells may possibly involve a 'catalytic' mechanism (1,10). Hexavalent Cr compounds are actively transported into cells, while trivalent Cr compounds enter cells ~ 1000 times less than hexavalent Cr (11). Following the entry of hexavalent Cr compounds, the chromium is reduced by various intracellular molecules, ultimately forming trivalent Cr. In die reduction process, oxygen radicals are likely to be formed and oxidative DNA damage may occur (12). Similarly, Ni 2+ can also undergo oxidation to Ni and subsequent reduction to Ni 2+ (13,14), however, these reactions occur less readily and to a lesser extent in biological systems compared with the reduction of chromate (13 — 15). Ni 2+ , in the presence of hydrogen peroxide and histidine, has been shown to generate oxidative damage in DNA (10,13,14). Thus, highly stable DNA-protein crosslinks induced by Ni are not likely to be due to the direct participation of Ni 2+ but are most likely due to crosslinks mediated indirectly (10,14). In the present study, a new approach was devised to study DNA—protein crosslinks present in intact cells. Cellular protein of cultured cells was radiolabeled with amino acids, DNA was isolated by standard proteinase K/phenol extraction, and the radioactive amino acids associated with purified DNA were measured. This approach allowed the detection of amino acid DNA complexes induced by Ni and Cr compounds and the study of their stability in intact cells.
X.Lin, Z.Zhuong and M.Costa 7
The sample was centrifuged at 12 000 g for 10 min, the pellet was washed with 70% ethancJ, centrifuged at 20 000 g for 10 min, the ethanol was decanted and the pellet was allowed to dry. The DNA was resuspended in 10 mM Tris or HEPES buffer (pH 8.0) until dissolved. DNA concentration was estimated by measuring the absorbance at 260 mm and 10 jig was taken to assess radioactivity using Aquasol counting medium and a Beckman scintillation counter. The specific activity of amino acids was determined in the cell suspension by assessing the radioactivity that was precipitated by 10% TCA added to the cell pellet and protein concentration was determined by use of a BCA assay (Sigma). This allowed the determination of counts per min per microgram of cellular protein for each radioactive amino acid and these values were used to adjust for differences in specific activity of protein labeling in each experiment. The stability of the amino acids complexed to DNA were studied using UltrafreeMC filter units (Millipore) or by precipitation of DNA as indicated in the respective figures and tables.
r
40
60
80
100
120
K,CrO4 Concentration (uM)
Table I. Comparison of the residual amino acid associated with DNA following potassium chromate treatment of intact CHO cells
Not*: •
P« 0.05;
" P< 0.01
Fig. 2. Comparison of the residual amino acids associated with DNA following potassium chromate treatment of intact cells. CHO cells were grown in media containing any one of the indicated radioactive amino acids for 24 h before treatment with 100 /iM of potassium chromate for 20 h. Residua] amino acids complexed with DNA were determined as described in the legend of Figure 1 and Materials and methods. Again, corrections were made for differences in specific activity of cellular proteins following potassium chromate treatment for each of these amino acids as outlined in Materials and methods and the legend of Figure 1. For Cys, Tyr, Thr, Ser, where there were increases, additional replicates were conducted to calculate the SEM and where indicated statistical significance (Student's /-test).
1764
Amino acid examined
Untreated
K2CrO4 (100 uM
Alanine Arginine Glycine Isoleucine Leucine Lysine Phenylalanine Proline Valine Serinc Threonine Tyrosine Tryptophane Aspartic acid Glutamic acid Methionine Histidine Cysteine
0.384° 2.375 1.795 0.253 0.469 1.219 1.858 1.752 0.531 57.670 96.740 3.410 1.998 0.592 2.207 0.095 5.641 0.038
0.495" 3.751 2.225 0.289 0.385 1.649 2.921 2.265 0.541 86.920 241.200 9.546 2.736 1.197 5.044 0.304 28.430 0.343
°pmol amino acid//ig DNA. Culture media deficient in the respective radioactive amino acid was utilized for labeling as described in Materials and methods. Following this labeling, cells were either left untreated or exposed to 100 jiM K2CrO4 for 20 h. DNA was isolated and the quantity of radioactive amino acid associated with DNA was determined (see Materials and methods). The actual quantity of unlabeled amino acids that may be present, intracellular and extracellular, was not measured and only the specific activity of the amino acid added was utilized for the estimations shown in the table. These calculations also assumed uniform labeling of cellular proteins but differences in the specific activity of proteins in chromate-treated compared to untreated cells was utilized as a correction factor. These values are a companion to those shown in Figure 2.
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
Fig. 1. Analysis of residual cysteine complexed with DNA following potassium chromate treatment. CHO cells were labeled with [33S]cysteine for 24 h prior to receiving 20 h exposure to the indicated concentrations of potassium chromate in complete a-minimal essential media (o-MEM). DNA was isolated by proteinase K/phenol/chloroform extraction as outlined in Materials and methods. Residual cysteine associated with DNA was determined by assessing radioactivity in each DNA sample. Since chromate caused an increase in radiolabeling as indicated by the specific activity of cellular proteins, the increased levels of specific activity were corrected for in each experiment as outlined in Materials and methods. This procedure involved analyzing the specific activity of cellular proteins and correcting the specific activity of cysteine associated with DNA by this factor. Each point is the mean ± SEM for three separate experiments. Starred points indicate P < 0.05 compared with appropriate control (Student's r-test).
Results Crosslinking of amino acids to DNA by chromate Figure 1 shows the dose-response relationship between the potassium chromate concentration and the crosslinking of residual cysteine to DNA. A detectable increase over control values in cysteine DNA crosslinks was observed at concentrations of 10 /iM, while at 50 and 100 fiM the crosslinking was statistically significant. Figure 2 examined the crosslinking of a number of amino acids to DNA by chromate. While cysteine was crosslinked to the DNA by chromate to a high level of statistical significance (Figure 2), hydroxyl containing amino acids such as Tyr and Thr were also significantly crosslinked to DNA as well as His or Met, which showed some increase. Table I (a companion to Figure 2) reports on estimated quantity of amino acids associated with DNA. These quantities were based upon estimations utilizing specific activities
DNA-protein crosslinks induced by Ni and Cr compounds
of the added amino acids, and there were no corrections for total endogenous levels, possible differences in compartmental labeling of proteins, etc., though corrections for the effect of chromate on protein-specific activity were made.
Cyitelne
Qlyclnc
Fig. 3. Effect of cyclohexamide on amino acid crosslinking to DNA. CHO cells were incubated for 4 h with cyclohexamide (40 fig/ml) while control cells received no cyclohexamide. Following this incubation radioactive amino acids were added to both control and cyclohexamide treated cultures for 4 h (20 /iCi/ml of each amino acid). The cells were washed with saline A and then selected cultures were incubated with potassium chromate (100 nM) for 5 h in a salt/glucose medium. Other cultures received no chromate but were also placed in salt/glucose media for 5 h. Following this incubation cells were collected by scraping and the level of amino acid associated with DNA was determined as described in Materials and methods.
The effect of nickel on crosslinking of amino acids to DNA in intact cells Figure 4 examines the complexing of [35S]cysteine to the DNA following treatment with various concentrations of NiCl2. NiCl2 (up to 0.5 mM) enhanced the cysteine complexing to the DNA by —1.6 times control levels. Concentrations beyond 0.5—1 mM did not further enhance the crosslinking of cysteine to DNA.
Table D. Stability of cysteine-DNA crosslink induced by K 2CrO4 in CHO cells* Treatment conditions
Untreated sample (%) [3H]DNAb' [33S]Cysteinec released released
K2CrO4-treated sample (%) [3H]DNA released
[35S]Cysteine released
>.
u EDTA 50 mM NaAc 25 mM (pH 4.0) 2-Mercaptoethanol (2.0%) EDTA 50 mM NaAc 25 mM (pH 4.0)
0.4 0.2
0 0
0.5 0
8.5 0.4
0.5
6.0
05
tw
0.4
3.7
L4
20.0
130-
•CHO cells were labeled with [35S]cysteine (10 /iCi/ml) and [3H]thymidine (0.1 /iCi/ml) for 24 h followed by incubation in fresh a-MEM with or without K2CrO4 (100 JJM) for 20 h. DNA was isolated by the method described in Materials and methods. [35S]cysteine and [3H]thymidine-labeled DNA-protein crosslink samples from untreated or K2CrO4-treated CHO cells were subjected to incubation for 3 h at 37°C under the conditions indicated in the table. The unbound amino acids and free nucleotides were separated by Ultrafree-MC filters (Millipore). ''Recovery of released 3H c.p.m. in the filtrate (percentage of original loading). c Recovery of released 35S c.p.m. in the filtrate (percentage of original loading).
10OU 0.5
10
15
2.0
2.5
NICI, cocentntlon ( mM ) Fig. 4. Residual cysteine complexed to DNA by NiCl2. CHO cells were labeled with [33S]cysteine for 24 h prior to receiving a 20 h exposure to the indicated concentrations of NiCl2. Residual cysteine associated with DNA was determined as described in the legend of Figure 1 and Materials and methods. Each value is the mean ± SEM for at least three determinations.
1765
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
o
o
It should be noted that attempts to degrade the DNA with proteolytic enzymes (peptidases) did not further reduce the amino acid radioactivity associated with DNA, suggesting that single amino acids were probably the major constituents of the crosslinked sample that had been studied (data not shown). Figure 3 examines the effect of cyclohexamide, a protein synthesis inhibitor, on the complexing of Cys, His and Gly to DNA. The data shows only the exposures to chromate compared with their respective controls. Cyclohexamide substantially increased chromate-induced cysteine—DNA crosslinks, suggesting that the cysteine came from glutathione and not protein-containing thiol. The stability of the cysteine DNA crosslinks induced by chromate were examined (Table II). Mercaptoethanol and EDTA enhanced the release of [35S]cysteine from chromate-induced crosslinks compared with those present as a background in untreated samples. EDTA had very little effect in untreated samples but increased the release of [35S]cysteine in chromatetreated samples. Acidic pH accelerated the ability of EDTA to release [35S]cysteine in chromate-treated samples. Since it was difficult to release completely all of the [35S]cysteine associated with the DNA in further washings using EDTA and mercaptoethanol, these results might suggest that there may be oxidatively induced rather than metal-mediated crosslinking of cysteine to DNA following chromate treatment.
X.Lin, Z.Zbuong and M.Costa
Discussion Studies of proteins complexed to DNA by chemical agents in intact cells have been complicated by the difficulty of stripping
the DNA of all background proteins, thus leaving only the crosslinked proteins for analysis (1). This problem of background has greatly limited the study of DNA-protein crosslinks. The present method utilizes isolation of DNA such that only very small residual radiolabeled peptides or amino acids were left associated with the DNA. In fact, further digestion of DNA with peptidase did not decrease the quantity of radioactive amino acid associated with DNA in these studies, suggesting that the predominant species associated with DNA were single amino acids. While it was still difficult to differentiate between metal catalyzed versus DNA-protein crosslinks with direct metal participation, it was clear that a portion of the chromate- or Ni-induced DNA-protein crosslinks were indirectly mediated because they could not be disrupted by treatment with EDTA. In the case of Ni 2+ , which weakly binds most ligands, the high stability of the amino acid complexed to DNA suggested that Ni 2+ did not participate directly in the formation of these complexes. In the case of chromate-induced complexes, it is possible that some Cr(III)mediated complexes were resistant to EDTA. Previous studies examining protein complexed to DNA with chromate demonstrated that EDTA could dissociate some of the DNA-protein complexes, suggesting that Cr(in) was in part mediating some DNA—protein crosslinks (15). However, not all the proteins associated with DNA could be extracted with extensive EDTA washing, suggesting that other bondings may be involved in the complexing of protein to DNA. It may also indicate that EDTA was not effective at dissociating all of the Cr(IH) bonds. The present study demonstrates that Ni did not seem to be directly involved in creating DNA—protein complexes since the small amount of Ni associated with DNA was easily released by EDTA, while 50% of the amino acid binding remained intact. In fact, the chemistry of Ni 2+ is such that most bonds with 200n
'H-Hiitldlnt
--Q--
170-
"S-Cyit.lnt
a c 5 c o
o
0.0
0.5
1.0
1.5
2.0
2 5
NICI, cocentratlon ( mM ) Time ( h )
Fig. 5. Residual histidine complexed with DNA induced by NiCl2 in CHO cells. CHO cells were labeled with [3H]histidine for 24 h prior to receiving 20 h exposure to the indicated concentration of NiClj. Residual hisudine associated with DNA was determined (see legend to Figure 1 and Materials and methods). Each value is the mean ± SEM for at least three determinants.
1766
Fig. 6. Time-dependent complexing of cysteine and histidine to DNA by NiCl2 in CHO cells. CHO cells were labeled with [35S)cysteine and [3H]histidine for 24 h prior to receiving exposure to 1 mM NiCl2 for the times indicated in the figure and the residual amino acids complexed to DNA were determined.
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
While there was a 1.8-fold increase in the crosslinking of histidine to DNA by NiCl2, the results were very similar to those found when cysteine was utilized (Figure 5). The time course for crosslinking of histidine and cysteine to DNA by NiCl2 indicated that longer incubations equalling a full cell cycle were required to produce DNA-protein crosslinks (Figure 6). Figure 7 examines the effect of hydrogen peroxide on cysteine and histidine crosslinking to DNA in intact cells. In these experiments, cells were labeled with both [3H]histidine and [ S]cysteine. NiCl2 alone produced an increase in the complexing of [3H]histidine and [^SJcysteine to DNA and hydrogen peroxide alone also produced some increase in the complexing of these amino acids to DNA. However, NiCl2 with hydrogen peroxide substantially enhanced the complexing of both histidine and cysteine to DNA. The enhancement was equivalent to the sum of the effect of NiCl2 and hydrogen peroxide alone. DMSO reduced some of the complexing of histidine and cysteine mediated by hydrogen peroxide and NiCl2 DUt did not lower the NiCl2-mediated crosslinking. The stability of the histidine and cysteine complexed to the DNA by NiCl2 was compared to the stability of ^NiflT) binding to DNA isolated from intact cells (Figure 8). As can be seen from the figure, the ^NitH) binding to the DNA was readily removed by EDTA washes, whereas 40—50% of the histidine or cysteine remained complexed to the DNA after these washes, suggesting that the complexing of these amino acids may not depend upon the ^NifTI) present, but may represent complexes mediated indirectly (i.e. catalyzed) by NiCl2.
DNA-protein crosslinks induced by Ni and Cr compounds
idized to Ni 3+ by H2O2 (13,14). This would result in the formation of hydroxyl radicals and other oxygen radicals, which could form malonyldialdehyde which crosslinks amino acids to DNA. DNA bases oxidized by oxygen radicals may also react with proteins directly to form DNA—protein crosslinks (10). Similarly, hexavalent chromate is reduced to various intermediate oxidation states (V,IV) and finally Cr(m), and this reduction has been shown to generate oxygen radicals which could indirectly crosslink protein to DNA (see above). The ability of DMSO, a • OH radical scavenger, to partly inhibit amino acid complexes to DNA induced by NiCl2 and H2O2 suggests that oxidative stress can be involved in creating histidine and cysteine DNA complexes. However, DMSO did not affect such complexes formed by NiCl2 alone. In the case of trivalent Cr, however, its binding is very stable. Thus, Cr(III) DNA-protein complexes are likely to survive SDS treatment as has been shown in previous studies (6). At the level of amino acid complexed to the DNA very similar results were observed in the present study as were found in previous work, demonstrating that Cr is directly participating in the complex as well as indirectly producing other complexes (1). It should be noted, however, that analysis of residual amino acids complexed to DNA cannot distinguish between complexing of amino acids or glutathione and complexes that involved entire proteins. In fact, chromium has been shown to complex glutathione and cysteine to DNA (16). Since cysteine residues are often buried in proteins, it is unlikely that they are readily accessible for chromium crosslinking and perhaps most of the cysteine —DNA crosslinks originated from glutathione. In fact the data in Figure 3 suggest that cysteine from glutathione is involved in chromate-induced DNA protein complexes. Therefore, analysis of histidine or tyrosine residues may repre-
I3H-Hlstldine l36 S-Cystelne
Trto n m n ( Ho.)
2 Fig. 7. Effect of H2O2 and DMSO on NiCI2-induced residual amino acid DNA complexes in CHO cells. CHO cells were labeled with [35S]cysteine and [3H]histidine before receiving a 20 h treatment of each of the following conditions: 1 mM NiCl2, 1 mM HJOJ, 1 mM NiCl2 and 1 mM H 2 Oj; 1 mM NiClj, 1 mM H2(>2 and 1 mM DMSO. Amino acids complexed with DNA were determined as described in the legend of Figure 1 and Materials and methods. Each value is the mean ± SEM for at least three determinants.
Trt*; EDTA m l m ( Ho.)
Fig. 8. Stability of NiCl2-induced amino acid DNA crosslink and "Ni DNA binding in CHO cells. CHO cells were labeled with [33S]cysteine and [3H]histidine 24 h prior to receiving exposure to 1 mM NiCl2 for 20 h (for amino acid DNA crosslink) or to 1 mM w NiCl 2 (20 /jCi/ml) for 20 h (for 63 Ni DNA binding). DNA samples were isolated as described in Materials and methods. Each DNA sample was incubated with 10 mM Tris, pH 8.0 (A) or 10 mM Tris and 20 mM EDTA, pH 8.0 (B) for 30 min at room temperature as indicated in the figure. Following each incubation, the sample was filtered through Ultrafree-MC filter units (Millipore) and the disassociated amino acid residues and w Ni radioactivity were determined.
1767
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
biological ligands were easily dissociated. Histidine and cysteine have the highest binding constants for Ni 2+ of any known small biological molecule (i.e. Cys = 4.4 X 108 M~', His 1.9 x 109 M~') (17). Binding constants for many other amino acids are much lower (105 M" 1 ) (17) while the binding constant of Ni(II) solubilized from Ni3S2 for DNA is very weak (7.3 X 102 M" 1 ) (18). In previous studies it was demonstrated that most of the proteins that were complexed to the DNA by Ni in intact cells were disrupted following treatment of DNA—protein complexes with SDS; however, DNA-protein complexes created by Ni could survive extractions with the less-stringent detergent sarcosyl (8). It was difficult to understand the contributions of Ni(II) on directly complexing amino acids to DNA; however, the amount of Ni ions associated with DNA was very small and was easily removed by washing even without EDTA, suggesting that NP+ binding to DNA was unlikely to be an important component in directly mediated amino acid DNA complexes. The fact that cysteine and histidine remained complexed to the DNA in Ni-treated cells through many washes with EDTA and SDS suggested that they were complexed indirectly, perhaps catalyzed by Ni 2+ /Ni redox cycling. Ni 2+ /Ni 3+ redox cycling is of interest because it is a minor reaction in biological systems and is specifically controlled by nickel binding to amino acids, peptides and proteins (13). Ni 2+ binds with a high affinity to the imidazole nitrogen of histidine and when bound at this site it can be ox-
X.Lin, Z.Zhuong and M.Costa
sent a more accurate picture of residual DNA-protein crosslinking. Acknowledgements The secretarial assistance of Jane Galvin is appreciated. This work was supported by grant nos ES 04895, ES 04715 and ES 05512 from the National Institute of Environmental Health Sciences and grant no. R814702 from the Environmental Protection Agency.
References
Received on April 7, 1992; revised on June 15, 1992; accepted on June 26, 1992
1768
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
1. Costa.M. (1991) DNA—protein complexes induced by chromate and other carcinogens. Environ. Health Persped., 92, 4 5 - 5 2 . 2. Patiemo.S.R. and Costa,M. (1985) DNA-ptotein crosslinks induced by nickel compounds in intact cultured mammalian cell. Chem.-Biol. Interactions, 55, 75-91. 3. Sugiyama.M., Wang,X.W. and Costa,M. (1986) Comparison of DNA lesions and cytotoxicity induced by calcium chromate in human, mouse and hamster cell lines. Cancer Res., 46, 4547-4551. 4. Sugiyama.M., Patiemo.S.R. and Costa.M. (1986) Characterization of DNA lesions induced by CaCrO4 in synchronous and asynchronous cultured mammalian cells. Mol. Pharmacol., 29, 606—613. 5. Miller,C.A.,m and Costa.M. (1989) Analysis of proteins crosslinked DNA after treatment of cells with formaldehyde, chromate and cis-diamminedichloroplatinum(II). Mol. Taricol., 2, 11-26. 6. Miller,C.A.,III and Costa.M. (1989) Immunological detection of DNA—protein complexes induced by chromate. Carcinogenesis, 10, 667-672. 7. Miller,C.A.,HI and Costa,M. (1990) Immunodetection of DNA-protein crosslinks by slot blotting. Mutat. Res., 23, 97-106. 8. Patierno.S.R. and Costa.M. (1987) Effects of nickel(II) on nuclear protein binding to DNA in intact mammalian cells. Cancer Biochem. Biophys., 9, 113-126. 9. Patierno.S.R., Sugiyama.M., BasilionJ.P. and Costa.M. (1985) Preferential DNA-protein crosslinking by NiCl2 in magnesium-insoluble regions of fractionated Chinese hamster ovary cell chromatin. Cancer Res., 45, 787—794. 10. Klein.C.B., Frenkel.K. and Costa.M. (1991) The role of oxidative processes in metal carcinogenesis. Chem. Res. Toxicol., 4, 592—604. 11. Arslan.P., Beltrame.M. and Tomasi.A. (1987) Intracellular chromium reduction. Biochim, Biophys. Ada, 931, 10-15. 12. Connett,P.H. and Wetterhahn.K. (1983) Metabolism of the carcinogen chromate by cellular constituents. Structure Bond., 54, 93 — 124. 13. Kasprzak.K.S. (1991) The role of oxidative damage in metal carcinogenicity. Chem. Res. Toxicol., 4, 604-615. 14. Datta.A.K., Misra.M., North.S.L. and Kasprzak.K. (1992) Enhancement of nickeKH) and L-histidine of 1 '-deoxyguanosine oxidation with hydrogen peroxide. Carcinogenesis, 13, 283-287. 15. Miller,C.A.,III and Costa.M. (1988) Characterization of DNA-protein completed induced in intact cells by the carcinogen chromate. Mol. Carcinogenesis, 1, 125-133. 16. Borges.K.M. and Wetterhahn.K.E. (1989) Chromium cross-links glutathione and cysteine to DNA. Carcinogenesis, 10, 2165-2168. 17. Stability Constants Supplement No. I, Special Publication 25. Chemical Society, London (1971). 18. LeeJ.E., Ciccarelli.R.B. and Jennette.K.W. (1982) Solubilization of the carcinogen nickel subsulfide and its interaction with deoxyribonucleic acid and protein. Biochemistry, 21, 771-778.
Analysis of residual amino acid—DNA crosslinks induced in intact cells by nickel and chromium compounds
Xinhua Lin, Zhbdong Zhnang and Max Costa1 New York University Medical Center, Nelson Institute of Environmental Medicine, New York, NY 10016, USA 'To whom all correspondence should be addressed
Introduction DNA-protein crosslinks represent an important lesion induced by chromium (Cr) and nickel (Ni) compounds in intact cells (1,2). These lesions are relatively persistent and not readily repaired (3,4). Previous studies have focused on examining the proteins complexed to DNA by Cr and Ni compounds in intact cells (2,5 - 7 ) . Cr has been shown to crosslink actin and other proteins found in the nuclear matrix while Ni compounds complexed both non-histone proteins and possibly histone HI (6,8). However, the complexing of proteins by Ni compounds in general has been found to be disrupted following treatment with SDS, while chromate-induced DNA—protein crosslinks were generally resistent to SDS treatment (5,9). Analysis of proteins complexed with DNA is plagued by problems of background inherent in DNA isolation procedures without proteases. Cr(in) was believed to participate in a substantial proportion of the Cr-induced DNA—protein crosslinks, whereas the binding of Ni 2+ to DNA and proteins was not as stable as the binding involving Cr(III) (6—9). Recent evidence has implicated both chromate and Ni compounds in the induction of oxidative DNA damage, and the possibility exists that the DNA-protein
Materials and methods Cell culture and amino acid labeling CHO cells were seeded at 4 x 10* cells in a 150 mm dish in 20 ml of complete medium containing 10% fetal bovine serum in an atmosphere of 5% CO}. After 4 h, radioactive amino acids (all were [3H] except cysonol meth [MS] or threonine [I4C]) were added at 10 jiCi/ml, and cells were incubated for another 24 h [cysteine 1105.7 (Ci/mmol), methionine 1162.9, tryptophane 17.9, glutarruc acid 25.0, aspartic acid 22.0, histidine 51.5, alanine 85.0, arginine 45.0, isoleucine 111.7, leucine60.0, rysine 85.7, phenylalanine 49.7, proline 35.8, valine 71.0, serine31.1, threonine 0.233, tyrosine 50.0]. Media deficient in each of the respective amino acids was utilized during labeling to enhance the sensitivity of detecting the amino acid-DNA complex. Cells were then washed twice with normal saline and 20 ml of complete a-MEM medium was added. Cells were exposed to various concentrations of NiC^ or K2CrO4 for 20 h as indicated in the figures and tables. Following this exposure, cells were scraped with a rubber policeman and collected by centrifugation at 1500 r.p.m. for 5 min. Cells were washed twice gently with PBS and resuspended in 10 ml of mis buffer. A sample of this suspension was withdrawn for determining the specific activity of the cellular proteins. The remaining cells were collected by centrifugation and placed in a digestion buffer containing 100 mM sodium chloride, 10 mM Tris, pH 8.0 and 0.5% SDS. This suspension was vortexed and allowed to stand to room temperature for 10 min. RNase (10 jig/ml) was added and the suspension was incubated at 37°C for 30 min. Proteinase K was subsequently added to a final concentration of 500 /jg/ml and the sample was digested at room temperature with gentle shaking overnight. The digested sample was thoroughly extracted twice with equal volume of phenol/chloroform/isoamyl alcohol (25:24:1) followed by another extraction with chloroform/isoamyl alcohol (24:1). The aqueous layer was transferred to new tubes and l/20th vol of saturated ammonium acetate and 2 vol of 100% ethanol were added and the mixture was allowed to stand at - 2 0 ° C for 1 h.
1763
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
Chinese hamster ovary cells were incubated with radioactive amino acids, the DNA was isolated by standard proteinase K/phenol/chJoroform extraction and residual amino acids complexed to the DNA were examined as an index of metal induced DNA-protein crosslinks. Using this method, both chromate and nickel caused residual histidine and cysteine to be complexed with the DNA isolated from metal-treated cells. In the case of chromate, a number of ammo acids were studied and Tyr, Thr and Cys were found to be complexed to DNA at a level (above the untreated control) that was statistically significant. Stability studies indicated that some of the chromate-induced DNA-protein complexes were mediated by direct participation of chromium (HI), whereas others that were resistent to dissociation by EDTA and mercaptoethanol did not seem to involve direct chromium(III) participation. A significant portion of the cysteine complexed to DNA by chromate was believed to involve glutathione since treatment of cells with cycloheximide did not decrease chromate-induced cysteine-DNA crosslinks. In the case of nickel, most of the stable DNA—protein crosslinks did not involve direct metal participation and were probably oxidatively mediated by Ni(II)/Ni(III) redox cycling. These findings present new methodology for analysis of DNA-protein crosslinks by examination of residual ammo acids associated with the DNA. This method should be highly sensitive and will yield important information about the mechanism of metal-induced DNA-protein crosslinks.
crosslinks induced by some of these compounds may be oxidatively mediated (10). Thus, while Cr(IH) may participate in some DNA-protein crosslinks, additional crosslinks induced by chromate in intact cells may possibly involve a 'catalytic' mechanism (1,10). Hexavalent Cr compounds are actively transported into cells, while trivalent Cr compounds enter cells ~ 1000 times less than hexavalent Cr (11). Following the entry of hexavalent Cr compounds, the chromium is reduced by various intracellular molecules, ultimately forming trivalent Cr. In die reduction process, oxygen radicals are likely to be formed and oxidative DNA damage may occur (12). Similarly, Ni 2+ can also undergo oxidation to Ni and subsequent reduction to Ni 2+ (13,14), however, these reactions occur less readily and to a lesser extent in biological systems compared with the reduction of chromate (13 — 15). Ni 2+ , in the presence of hydrogen peroxide and histidine, has been shown to generate oxidative damage in DNA (10,13,14). Thus, highly stable DNA-protein crosslinks induced by Ni are not likely to be due to the direct participation of Ni 2+ but are most likely due to crosslinks mediated indirectly (10,14). In the present study, a new approach was devised to study DNA—protein crosslinks present in intact cells. Cellular protein of cultured cells was radiolabeled with amino acids, DNA was isolated by standard proteinase K/phenol extraction, and the radioactive amino acids associated with purified DNA were measured. This approach allowed the detection of amino acid DNA complexes induced by Ni and Cr compounds and the study of their stability in intact cells.
X.Lin, Z.Zhuong and M.Costa 7
The sample was centrifuged at 12 000 g for 10 min, the pellet was washed with 70% ethancJ, centrifuged at 20 000 g for 10 min, the ethanol was decanted and the pellet was allowed to dry. The DNA was resuspended in 10 mM Tris or HEPES buffer (pH 8.0) until dissolved. DNA concentration was estimated by measuring the absorbance at 260 mm and 10 jig was taken to assess radioactivity using Aquasol counting medium and a Beckman scintillation counter. The specific activity of amino acids was determined in the cell suspension by assessing the radioactivity that was precipitated by 10% TCA added to the cell pellet and protein concentration was determined by use of a BCA assay (Sigma). This allowed the determination of counts per min per microgram of cellular protein for each radioactive amino acid and these values were used to adjust for differences in specific activity of protein labeling in each experiment. The stability of the amino acids complexed to DNA were studied using UltrafreeMC filter units (Millipore) or by precipitation of DNA as indicated in the respective figures and tables.
r
40
60
80
100
120
K,CrO4 Concentration (uM)
Table I. Comparison of the residual amino acid associated with DNA following potassium chromate treatment of intact CHO cells
Not*: •
P« 0.05;
" P< 0.01
Fig. 2. Comparison of the residual amino acids associated with DNA following potassium chromate treatment of intact cells. CHO cells were grown in media containing any one of the indicated radioactive amino acids for 24 h before treatment with 100 /iM of potassium chromate for 20 h. Residua] amino acids complexed with DNA were determined as described in the legend of Figure 1 and Materials and methods. Again, corrections were made for differences in specific activity of cellular proteins following potassium chromate treatment for each of these amino acids as outlined in Materials and methods and the legend of Figure 1. For Cys, Tyr, Thr, Ser, where there were increases, additional replicates were conducted to calculate the SEM and where indicated statistical significance (Student's /-test).
1764
Amino acid examined
Untreated
K2CrO4 (100 uM
Alanine Arginine Glycine Isoleucine Leucine Lysine Phenylalanine Proline Valine Serinc Threonine Tyrosine Tryptophane Aspartic acid Glutamic acid Methionine Histidine Cysteine
0.384° 2.375 1.795 0.253 0.469 1.219 1.858 1.752 0.531 57.670 96.740 3.410 1.998 0.592 2.207 0.095 5.641 0.038
0.495" 3.751 2.225 0.289 0.385 1.649 2.921 2.265 0.541 86.920 241.200 9.546 2.736 1.197 5.044 0.304 28.430 0.343
°pmol amino acid//ig DNA. Culture media deficient in the respective radioactive amino acid was utilized for labeling as described in Materials and methods. Following this labeling, cells were either left untreated or exposed to 100 jiM K2CrO4 for 20 h. DNA was isolated and the quantity of radioactive amino acid associated with DNA was determined (see Materials and methods). The actual quantity of unlabeled amino acids that may be present, intracellular and extracellular, was not measured and only the specific activity of the amino acid added was utilized for the estimations shown in the table. These calculations also assumed uniform labeling of cellular proteins but differences in the specific activity of proteins in chromate-treated compared to untreated cells was utilized as a correction factor. These values are a companion to those shown in Figure 2.
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
Fig. 1. Analysis of residual cysteine complexed with DNA following potassium chromate treatment. CHO cells were labeled with [33S]cysteine for 24 h prior to receiving 20 h exposure to the indicated concentrations of potassium chromate in complete a-minimal essential media (o-MEM). DNA was isolated by proteinase K/phenol/chloroform extraction as outlined in Materials and methods. Residual cysteine associated with DNA was determined by assessing radioactivity in each DNA sample. Since chromate caused an increase in radiolabeling as indicated by the specific activity of cellular proteins, the increased levels of specific activity were corrected for in each experiment as outlined in Materials and methods. This procedure involved analyzing the specific activity of cellular proteins and correcting the specific activity of cysteine associated with DNA by this factor. Each point is the mean ± SEM for three separate experiments. Starred points indicate P < 0.05 compared with appropriate control (Student's r-test).
Results Crosslinking of amino acids to DNA by chromate Figure 1 shows the dose-response relationship between the potassium chromate concentration and the crosslinking of residual cysteine to DNA. A detectable increase over control values in cysteine DNA crosslinks was observed at concentrations of 10 /iM, while at 50 and 100 fiM the crosslinking was statistically significant. Figure 2 examined the crosslinking of a number of amino acids to DNA by chromate. While cysteine was crosslinked to the DNA by chromate to a high level of statistical significance (Figure 2), hydroxyl containing amino acids such as Tyr and Thr were also significantly crosslinked to DNA as well as His or Met, which showed some increase. Table I (a companion to Figure 2) reports on estimated quantity of amino acids associated with DNA. These quantities were based upon estimations utilizing specific activities
DNA-protein crosslinks induced by Ni and Cr compounds
of the added amino acids, and there were no corrections for total endogenous levels, possible differences in compartmental labeling of proteins, etc., though corrections for the effect of chromate on protein-specific activity were made.
Cyitelne
Qlyclnc
Fig. 3. Effect of cyclohexamide on amino acid crosslinking to DNA. CHO cells were incubated for 4 h with cyclohexamide (40 fig/ml) while control cells received no cyclohexamide. Following this incubation radioactive amino acids were added to both control and cyclohexamide treated cultures for 4 h (20 /iCi/ml of each amino acid). The cells were washed with saline A and then selected cultures were incubated with potassium chromate (100 nM) for 5 h in a salt/glucose medium. Other cultures received no chromate but were also placed in salt/glucose media for 5 h. Following this incubation cells were collected by scraping and the level of amino acid associated with DNA was determined as described in Materials and methods.
The effect of nickel on crosslinking of amino acids to DNA in intact cells Figure 4 examines the complexing of [35S]cysteine to the DNA following treatment with various concentrations of NiCl2. NiCl2 (up to 0.5 mM) enhanced the cysteine complexing to the DNA by —1.6 times control levels. Concentrations beyond 0.5—1 mM did not further enhance the crosslinking of cysteine to DNA.
Table D. Stability of cysteine-DNA crosslink induced by K 2CrO4 in CHO cells* Treatment conditions
Untreated sample (%) [3H]DNAb' [33S]Cysteinec released released
K2CrO4-treated sample (%) [3H]DNA released
[35S]Cysteine released
>.
u EDTA 50 mM NaAc 25 mM (pH 4.0) 2-Mercaptoethanol (2.0%) EDTA 50 mM NaAc 25 mM (pH 4.0)
0.4 0.2
0 0
0.5 0
8.5 0.4
0.5
6.0
05
tw
0.4
3.7
L4
20.0
130-
•CHO cells were labeled with [35S]cysteine (10 /iCi/ml) and [3H]thymidine (0.1 /iCi/ml) for 24 h followed by incubation in fresh a-MEM with or without K2CrO4 (100 JJM) for 20 h. DNA was isolated by the method described in Materials and methods. [35S]cysteine and [3H]thymidine-labeled DNA-protein crosslink samples from untreated or K2CrO4-treated CHO cells were subjected to incubation for 3 h at 37°C under the conditions indicated in the table. The unbound amino acids and free nucleotides were separated by Ultrafree-MC filters (Millipore). ''Recovery of released 3H c.p.m. in the filtrate (percentage of original loading). c Recovery of released 35S c.p.m. in the filtrate (percentage of original loading).
10OU 0.5
10
15
2.0
2.5
NICI, cocentntlon ( mM ) Fig. 4. Residual cysteine complexed to DNA by NiCl2. CHO cells were labeled with [33S]cysteine for 24 h prior to receiving a 20 h exposure to the indicated concentrations of NiCl2. Residual cysteine associated with DNA was determined as described in the legend of Figure 1 and Materials and methods. Each value is the mean ± SEM for at least three determinations.
1765
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
o
o
It should be noted that attempts to degrade the DNA with proteolytic enzymes (peptidases) did not further reduce the amino acid radioactivity associated with DNA, suggesting that single amino acids were probably the major constituents of the crosslinked sample that had been studied (data not shown). Figure 3 examines the effect of cyclohexamide, a protein synthesis inhibitor, on the complexing of Cys, His and Gly to DNA. The data shows only the exposures to chromate compared with their respective controls. Cyclohexamide substantially increased chromate-induced cysteine—DNA crosslinks, suggesting that the cysteine came from glutathione and not protein-containing thiol. The stability of the cysteine DNA crosslinks induced by chromate were examined (Table II). Mercaptoethanol and EDTA enhanced the release of [35S]cysteine from chromate-induced crosslinks compared with those present as a background in untreated samples. EDTA had very little effect in untreated samples but increased the release of [35S]cysteine in chromatetreated samples. Acidic pH accelerated the ability of EDTA to release [35S]cysteine in chromate-treated samples. Since it was difficult to release completely all of the [35S]cysteine associated with the DNA in further washings using EDTA and mercaptoethanol, these results might suggest that there may be oxidatively induced rather than metal-mediated crosslinking of cysteine to DNA following chromate treatment.
X.Lin, Z.Zbuong and M.Costa
Discussion Studies of proteins complexed to DNA by chemical agents in intact cells have been complicated by the difficulty of stripping
the DNA of all background proteins, thus leaving only the crosslinked proteins for analysis (1). This problem of background has greatly limited the study of DNA-protein crosslinks. The present method utilizes isolation of DNA such that only very small residual radiolabeled peptides or amino acids were left associated with the DNA. In fact, further digestion of DNA with peptidase did not decrease the quantity of radioactive amino acid associated with DNA in these studies, suggesting that the predominant species associated with DNA were single amino acids. While it was still difficult to differentiate between metal catalyzed versus DNA-protein crosslinks with direct metal participation, it was clear that a portion of the chromate- or Ni-induced DNA-protein crosslinks were indirectly mediated because they could not be disrupted by treatment with EDTA. In the case of Ni 2+ , which weakly binds most ligands, the high stability of the amino acid complexed to DNA suggested that Ni 2+ did not participate directly in the formation of these complexes. In the case of chromate-induced complexes, it is possible that some Cr(III)mediated complexes were resistant to EDTA. Previous studies examining protein complexed to DNA with chromate demonstrated that EDTA could dissociate some of the DNA-protein complexes, suggesting that Cr(in) was in part mediating some DNA—protein crosslinks (15). However, not all the proteins associated with DNA could be extracted with extensive EDTA washing, suggesting that other bondings may be involved in the complexing of protein to DNA. It may also indicate that EDTA was not effective at dissociating all of the Cr(IH) bonds. The present study demonstrates that Ni did not seem to be directly involved in creating DNA—protein complexes since the small amount of Ni associated with DNA was easily released by EDTA, while 50% of the amino acid binding remained intact. In fact, the chemistry of Ni 2+ is such that most bonds with 200n
'H-Hiitldlnt
--Q--
170-
"S-Cyit.lnt
a c 5 c o
o
0.0
0.5
1.0
1.5
2.0
2 5
NICI, cocentratlon ( mM ) Time ( h )
Fig. 5. Residual histidine complexed with DNA induced by NiCl2 in CHO cells. CHO cells were labeled with [3H]histidine for 24 h prior to receiving 20 h exposure to the indicated concentration of NiClj. Residual hisudine associated with DNA was determined (see legend to Figure 1 and Materials and methods). Each value is the mean ± SEM for at least three determinants.
1766
Fig. 6. Time-dependent complexing of cysteine and histidine to DNA by NiCl2 in CHO cells. CHO cells were labeled with [35S)cysteine and [3H]histidine for 24 h prior to receiving exposure to 1 mM NiCl2 for the times indicated in the figure and the residual amino acids complexed to DNA were determined.
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
While there was a 1.8-fold increase in the crosslinking of histidine to DNA by NiCl2, the results were very similar to those found when cysteine was utilized (Figure 5). The time course for crosslinking of histidine and cysteine to DNA by NiCl2 indicated that longer incubations equalling a full cell cycle were required to produce DNA-protein crosslinks (Figure 6). Figure 7 examines the effect of hydrogen peroxide on cysteine and histidine crosslinking to DNA in intact cells. In these experiments, cells were labeled with both [3H]histidine and [ S]cysteine. NiCl2 alone produced an increase in the complexing of [3H]histidine and [^SJcysteine to DNA and hydrogen peroxide alone also produced some increase in the complexing of these amino acids to DNA. However, NiCl2 with hydrogen peroxide substantially enhanced the complexing of both histidine and cysteine to DNA. The enhancement was equivalent to the sum of the effect of NiCl2 and hydrogen peroxide alone. DMSO reduced some of the complexing of histidine and cysteine mediated by hydrogen peroxide and NiCl2 DUt did not lower the NiCl2-mediated crosslinking. The stability of the histidine and cysteine complexed to the DNA by NiCl2 was compared to the stability of ^NiflT) binding to DNA isolated from intact cells (Figure 8). As can be seen from the figure, the ^NitH) binding to the DNA was readily removed by EDTA washes, whereas 40—50% of the histidine or cysteine remained complexed to the DNA after these washes, suggesting that the complexing of these amino acids may not depend upon the ^NifTI) present, but may represent complexes mediated indirectly (i.e. catalyzed) by NiCl2.
DNA-protein crosslinks induced by Ni and Cr compounds
idized to Ni 3+ by H2O2 (13,14). This would result in the formation of hydroxyl radicals and other oxygen radicals, which could form malonyldialdehyde which crosslinks amino acids to DNA. DNA bases oxidized by oxygen radicals may also react with proteins directly to form DNA—protein crosslinks (10). Similarly, hexavalent chromate is reduced to various intermediate oxidation states (V,IV) and finally Cr(m), and this reduction has been shown to generate oxygen radicals which could indirectly crosslink protein to DNA (see above). The ability of DMSO, a • OH radical scavenger, to partly inhibit amino acid complexes to DNA induced by NiCl2 and H2O2 suggests that oxidative stress can be involved in creating histidine and cysteine DNA complexes. However, DMSO did not affect such complexes formed by NiCl2 alone. In the case of trivalent Cr, however, its binding is very stable. Thus, Cr(III) DNA-protein complexes are likely to survive SDS treatment as has been shown in previous studies (6). At the level of amino acid complexed to the DNA very similar results were observed in the present study as were found in previous work, demonstrating that Cr is directly participating in the complex as well as indirectly producing other complexes (1). It should be noted, however, that analysis of residual amino acids complexed to DNA cannot distinguish between complexing of amino acids or glutathione and complexes that involved entire proteins. In fact, chromium has been shown to complex glutathione and cysteine to DNA (16). Since cysteine residues are often buried in proteins, it is unlikely that they are readily accessible for chromium crosslinking and perhaps most of the cysteine —DNA crosslinks originated from glutathione. In fact the data in Figure 3 suggest that cysteine from glutathione is involved in chromate-induced DNA protein complexes. Therefore, analysis of histidine or tyrosine residues may repre-
I3H-Hlstldine l36 S-Cystelne
Trto n m n ( Ho.)
2 Fig. 7. Effect of H2O2 and DMSO on NiCI2-induced residual amino acid DNA complexes in CHO cells. CHO cells were labeled with [35S]cysteine and [3H]histidine before receiving a 20 h treatment of each of the following conditions: 1 mM NiCl2, 1 mM HJOJ, 1 mM NiCl2 and 1 mM H 2 Oj; 1 mM NiClj, 1 mM H2(>2 and 1 mM DMSO. Amino acids complexed with DNA were determined as described in the legend of Figure 1 and Materials and methods. Each value is the mean ± SEM for at least three determinants.
Trt*; EDTA m l m ( Ho.)
Fig. 8. Stability of NiCl2-induced amino acid DNA crosslink and "Ni DNA binding in CHO cells. CHO cells were labeled with [33S]cysteine and [3H]histidine 24 h prior to receiving exposure to 1 mM NiCl2 for 20 h (for amino acid DNA crosslink) or to 1 mM w NiCl 2 (20 /jCi/ml) for 20 h (for 63 Ni DNA binding). DNA samples were isolated as described in Materials and methods. Each DNA sample was incubated with 10 mM Tris, pH 8.0 (A) or 10 mM Tris and 20 mM EDTA, pH 8.0 (B) for 30 min at room temperature as indicated in the figure. Following each incubation, the sample was filtered through Ultrafree-MC filter units (Millipore) and the disassociated amino acid residues and w Ni radioactivity were determined.
1767
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
biological ligands were easily dissociated. Histidine and cysteine have the highest binding constants for Ni 2+ of any known small biological molecule (i.e. Cys = 4.4 X 108 M~', His 1.9 x 109 M~') (17). Binding constants for many other amino acids are much lower (105 M" 1 ) (17) while the binding constant of Ni(II) solubilized from Ni3S2 for DNA is very weak (7.3 X 102 M" 1 ) (18). In previous studies it was demonstrated that most of the proteins that were complexed to the DNA by Ni in intact cells were disrupted following treatment of DNA—protein complexes with SDS; however, DNA-protein complexes created by Ni could survive extractions with the less-stringent detergent sarcosyl (8). It was difficult to understand the contributions of Ni(II) on directly complexing amino acids to DNA; however, the amount of Ni ions associated with DNA was very small and was easily removed by washing even without EDTA, suggesting that NP+ binding to DNA was unlikely to be an important component in directly mediated amino acid DNA complexes. The fact that cysteine and histidine remained complexed to the DNA in Ni-treated cells through many washes with EDTA and SDS suggested that they were complexed indirectly, perhaps catalyzed by Ni 2+ /Ni redox cycling. Ni 2+ /Ni 3+ redox cycling is of interest because it is a minor reaction in biological systems and is specifically controlled by nickel binding to amino acids, peptides and proteins (13). Ni 2+ binds with a high affinity to the imidazole nitrogen of histidine and when bound at this site it can be ox-
X.Lin, Z.Zhuong and M.Costa
sent a more accurate picture of residual DNA-protein crosslinking. Acknowledgements The secretarial assistance of Jane Galvin is appreciated. This work was supported by grant nos ES 04895, ES 04715 and ES 05512 from the National Institute of Environmental Health Sciences and grant no. R814702 from the Environmental Protection Agency.
References
Received on April 7, 1992; revised on June 15, 1992; accepted on June 26, 1992
1768
Downloaded from http://carcin.oxfordjournals.org/ at Johns Hopkins University on June 8, 2015
1. Costa.M. (1991) DNA—protein complexes induced by chromate and other carcinogens. Environ. Health Persped., 92, 4 5 - 5 2 . 2. Patiemo.S.R. and Costa,M. (1985) DNA-ptotein crosslinks induced by nickel compounds in intact cultured mammalian cell. Chem.-Biol. Interactions, 55, 75-91. 3. Sugiyama.M., Wang,X.W. and Costa,M. (1986) Comparison of DNA lesions and cytotoxicity induced by calcium chromate in human, mouse and hamster cell lines. Cancer Res., 46, 4547-4551. 4. Sugiyama.M., Patiemo.S.R. and Costa.M. (1986) Characterization of DNA lesions induced by CaCrO4 in synchronous and asynchronous cultured mammalian cells. Mol. Pharmacol., 29, 606—613. 5. Miller,C.A.,m and Costa.M. (1989) Analysis of proteins crosslinked DNA after treatment of cells with formaldehyde, chromate and cis-diamminedichloroplatinum(II). Mol. Taricol., 2, 11-26. 6. Miller,C.A.,III and Costa.M. (1989) Immunological detection of DNA—protein complexes induced by chromate. Carcinogenesis, 10, 667-672. 7. Miller,C.A.,HI and Costa,M. (1990) Immunodetection of DNA-protein crosslinks by slot blotting. Mutat. Res., 23, 97-106. 8. Patierno.S.R. and Costa.M. (1987) Effects of nickel(II) on nuclear protein binding to DNA in intact mammalian cells. Cancer Biochem. Biophys., 9, 113-126. 9. Patierno.S.R., Sugiyama.M., BasilionJ.P. and Costa.M. (1985) Preferential DNA-protein crosslinking by NiCl2 in magnesium-insoluble regions of fractionated Chinese hamster ovary cell chromatin. Cancer Res., 45, 787—794. 10. Klein.C.B., Frenkel.K. and Costa.M. (1991) The role of oxidative processes in metal carcinogenesis. Chem. Res. Toxicol., 4, 592—604. 11. Arslan.P., Beltrame.M. and Tomasi.A. (1987) Intracellular chromium reduction. Biochim, Biophys. Ada, 931, 10-15. 12. Connett,P.H. and Wetterhahn.K. (1983) Metabolism of the carcinogen chromate by cellular constituents. Structure Bond., 54, 93 — 124. 13. Kasprzak.K.S. (1991) The role of oxidative damage in metal carcinogenicity. Chem. Res. Toxicol., 4, 604-615. 14. Datta.A.K., Misra.M., North.S.L. and Kasprzak.K. (1992) Enhancement of nickeKH) and L-histidine of 1 '-deoxyguanosine oxidation with hydrogen peroxide. Carcinogenesis, 13, 283-287. 15. Miller,C.A.,III and Costa.M. (1988) Characterization of DNA-protein completed induced in intact cells by the carcinogen chromate. Mol. Carcinogenesis, 1, 125-133. 16. Borges.K.M. and Wetterhahn.K.E. (1989) Chromium cross-links glutathione and cysteine to DNA. Carcinogenesis, 10, 2165-2168. 17. Stability Constants Supplement No. I, Special Publication 25. Chemical Society, London (1971). 18. LeeJ.E., Ciccarelli.R.B. and Jennette.K.W. (1982) Solubilization of the carcinogen nickel subsulfide and its interaction with deoxyribonucleic acid and protein. Biochemistry, 21, 771-778.
